the diversity pf dual eligible beneficiaries: an ... · pdf filemedicaid kaiser i commission...

medicaid

kaiser commiss ion o nI

S

S

U

E

P

A

P

E

R

a n d t h e uninsured

1 3 3 0 G S T R E E T N W , W A S H I N G T O N , D C 2 0 0 0 5P H O N E : 2 0 2 - 3 4 7 - 5 2 7 0 , F A X : 2 0 2 - 3 4 7 - 5 2 7 4W E B S I T E : W W W . K F F . O R G

April 2012

The Diversity of Dual Eligible Beneficiaries: An Examination of Services and Spending for People

Eligible for Both Medicaid and Medicare

Nationwide, 9 million individuals are dually eligible for both Medicaid and Medicare services. These “dual eligibles” are low-income seniors and individuals with disabilities who rely on Medicare for coverage of acute care medical services (such as hospital, physician, prescription drugs, and post-acute care) and on Medicaid for financial assistance with Medicare’s premiums and cost sharing. Most dual eligibles also rely on Medicaid to provide coverage for services not included in Medicare, particularly long-term care. Dual eligibles have long been of interest to policymakers due to their relatively high health care needs and correspondingly high cost: Although dual eligibles represent only 21 percent of the Medicare population and 15 percent of the Medicaid population, they account for nearly 36 percent of total Medicare spending and 39 percent of total Medicaid spending.1

Policymakers are exploring strategies to better coordinate and integrate care for dual eligibles and align financing for this population.2 Finding realistic, effective strategies is challenging for many reasons. Medicare and Medicaid are two different, very large public health insurance programs that operate separately and sometimes work at cross-purposes. Medicare is a federal program while Medicaid is a joint federal-state program that differs across states. Further, there is great diversity among dual eligibles in their health care needs. Some dual eligibles have relatively limited needs whereas others are among the nation’s most vulnerable individuals with complex health conditions and high medical and long-term care costs.

Despite their considerable policy importance, limited work has been done that examines combined Medicaid and Medicare service use and spending patterns for dual eligibles. In this brief, we update a previous study3 and present findings based on analysis of linked 2007 Medicare and Medicaid data. Specifically, we examine characteristics, health status, utilization, and spending for dual eligibles compared to the non-dual Medicare population.4

DUAL ELIGIBLES ARE A DIVERSE POPULATION

There is considerable diversity within the dual eligible population. First, there are different eligibility pathways through which a Medicare beneficiary may become eligible for Medicaid, depending on the person’s income and assets.5 Most dual eligibles (76.3 percent or about 6.8 million individuals) are “full

002

dual eligibles” and are eligible for all Medicaid benefits that provide “wrap-around coverage” to Medicare (Table 1). This wrap-around coverage includes some very significant categories of benefits for dual eligibles that Medicare does not cover, such as long-term care. The extent of this wrap-around coverage depends upon what services are provided by the state Medicaid program where a dual eligible lives. In addition to receiving all Medicaid benefits, Medicaid also pays Medicare cost sharing for full dual eligibles. The balance of dual eligibles (22.5 percent or about 2.0 million individuals) is comprised of so-called “partial dual eligibles.” Sometimes referred to as Medicare Savings Plan or MSP dual eligibles, partial dual eligibles are not eligible for full Medicaid benefits and only receive help with Medicare cost sharing from Medicaid.

Dual eligibles also vary on other dimensions. There is wide variation in functional status among dual eligibles. A quarter (25.1 percent) report limitations in three or more activities of daily living (ADLs), but the bulk (56.7 percent) report no such limitations. Similarly, many dual eligibles have serious chronic health conditions (e.g., diabetes, mental illness and heart disease), but nearly a third (31.9 percent) had no health conditions among the ones we examined (data not shown).6 Though a sizable minority (17.0 percent) of dual eligibles lives in institutions, most (79.2 percent) live in the community; another 3.8 percent spend time in both the community and in institutions.

There is considerable age difference among dual eligibles. About one-third of the group is disabled and under age 65, but two-thirds are over age 65 (Table 1). Dual eligibles who are under age 65 differ from those age 65 and up (Figure 1). For example, elderly dual eligibles are more likely than non-elderly dual eligibles to qualify as MSP or partial (versus full) dual eligibles, more likely to be female, and more likely to be married (versus single, widowed or divorced). Elderly dual eligibles are also more likely to have a functional limitation and, correspondingly, more likely to reside in an institutional setting.

Figure 1

Selected Characteristics of Medicaid-Medicare Dual Eligibles by Age, 2007

21.7

49.1

23.2

58.3

71.7

73.6

9.3

33.5

15.8

61.7

51.9

80.6

Live in institution

1+ ADLs

Married

White Non-Hispanic

Female

Full Medicaid benefits

Duals Age <65 Duals Age 65+

SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured.* (**) (***) Non-elderly dual beneficiaries are significantly different from elderly dual beneficiaries at the .10 (.05) (.01) level, two-tailed test.

***

***

***

***

***

00 3

DUAL ELIGIBLES ARE DISTINCT FROM OTHER MEDICARE BENEFICIARIES

Apart from being a diverse group in and of itself, dual eligibles as a group differ from other Medicare beneficiaries (Figure 2). For example, dual eligibles are poorer. More than half (53.4 percent) have incomes less than $10,000 whereas only 8.3 percent of other Medicare beneficiaries have incomes at this level. Dual eligibles are younger: 37.5 percent of dual eligibles are under age 65 versus 11.9 percent for other Medicare beneficiaries. They are also less likely to be married and to be of a race/ethnicity other than White non-Hispanic. Finally, and importantly, dual eligibles are much more likely to be living in an institution—one out of every six dual eligibles lives in an institution whereas only one out of 50 other Medicare beneficiaries reside in an institution. This comparatively high rate of institutionalization among dual eligibles has important implications on their health care spending.

Dual eligibles are also in poorer health. As shown in Figure 3, compared to other Medicare beneficiaries, dual eligibles are much more likely to be diagnosed with diabetes (5.2 percent versus 0.5 percent), COPD/lung disease (25.1 percent versus 16.3 percent), mental illness (34.0 percent versus 16.8 percent) and Alzheimer’s (5.7 percent versus 2.0 percent), among others. Dual eligibles also have a lower functional status. For example, more than a quarter of dual eligibles (25.1 percent) are limited in three or more activities of daily living (ADLs) whereas only 6.3 percent of other Medicare beneficiaries fall into this category (Table 1).

Figure 2

Selected Characteristics of Medicaid-Medicare Dual Eligibles and Other Medicare Beneficiaries, 2007

1.9

53.7

83.2

57.9

11.9

8.3

17.0

64.3

59.5

20.4

37.5

53.4

Live in institution

Female

White non-Hispanic

Married

Age <65 years

Income ≤ $10,000

Medicaid-Medicare Dual Beneficiaries Other Medicare Beneficiaries

SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured.* (**) (***) Medicaid-Medicare dual beneficiaries are significantly different from other Medicare beneficiaries at the .10 (.05) (.01) level, two-tailed test.

***

***

***

*** ***

***

*** ***

Figure 3

Health Status of Medicaid-Medicare Dual Eligibles and Other Medicare Beneficiaries, 2007

25.6

0.6

2.0

16.8

16.3

0.5

29.3

6.3

5.7

34.0

25.1

5.2

Heart Disease

Mental Retardation

Alzheimers

Mental Illness

COPD/Lung disease

Diabetes


SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured.* (**) (***) Medicaid-Medicare dual beneficiaries are significantly different from other Medicare beneficiaries at the .10 (.05) (.01) level, two-tailed test.

***

***

***

***

***

***

***

***

004

Figure 4

Service Use Among Medicaid-Medicare Dual Eligibles and Other Medicare Beneficiaries, 2007

62.8

3.5

0

15.1

51.2

65.8

9.2

16.0

26.1

66.9

Physician Visit

Skilled Nursing Facillity

Institutional Long-Term Care

Inpatient Hospital

Outpatient Hospital


*

SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured.NOTE: Includes only services paid for by Medicare (non-duals) and Medicare or Medicaid (duals).* (**) (***) Medicaid-Medicare dual beneficiaries are significantly different from other Medicare beneficiaries at the .10 (.05) (.01) level, two-tailed test.

***

***

***

***

***

DUAL ELIGIBLES USE MORE HEALTH CARE SERVICES COMPARED TO OTHER MEDICARE

BENEFICIARIES AND COST MORE TO SERVE

Given their lower health status, dual eligibles have a higher level of service use in all of the categories examined compared to other Medicare beneficiaries (Figure 4).7 These differences are particularly large for outpatient hospital services, inpatient hospital services and institutional long-term care. Among dual eligibles, there are significant utilization differences between elderly and non-elderly dual eligibles (Figure 5). Elderly dual eligibles are much more likely to use institutional long-term care and skilled nursing facilities compared to their younger counterparts, while non-elderly dual eligibles are more likely to use outpatient hospital and physician services than their elderly counterparts. For inpatient hospital care, however, we found no use difference between elderly and nonelderly dual eligibles.

Figure 5

Service Use Among Medicaid-Medicare Dual Eligibles by Age, 2007

62.0

64.0

27.5

11.9

21.6

72.2

71.8

23.9

4.5

6.7

Physician Visit

Outpatient Hospital

Inpatient Hospital

Skilled Nursing Facillity

Institutional Long-Term Care

Dual Age <65 Dual Age 65+

SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured.NOTE: Includes only services paid for by Medicare or Medicaid.* (**) (***) Non-elderly dual beneficiaries are significantly different from elderly dual beneficiaries at the .10 (.05) (.01) level, two-tailed test.

***

***

***

***

00 5

Corresponding to their higher service use, dual eligibles are more costly to care for (Figure 6). In 2007, Medicare alone spent $15,850 per capita on dual eligbiles, more than twice the per capita spending on other Medicare beneficiaries ($7,226). Accounting for both federal and state Medicaid spending, total Medicaid per capita spending on dual eligibles is $14,018, a bit lower than the per capita Medicare spending amount. Combined per capita Medicare and Medicaid spending for dual eligibles is $29,868, more than four times per capita spending for other Medicare beneficiaries. Among dual eligibles, elderly are slightly more costly to care for than non-elderly dual eligibles. Combined Medicare and Medicaid spending per capita for the two groups were, respectively, $31,139 and $27,747 in 2007 (Figure 7).

In terms of total spending (Figure 8), Medicare and Medicaid spending for the 8.9 million dual eligibles totaled $265.7 billion in 2007. For the 35.4 million other

Medicare beneficiaries (a group about four times larger than the dual eligible group), Medicare spending totaled $256.0 billion.

Medicare and Medicaid finance nearly equal shares of spending for dual eligibles: $141.0 billion and $124.7 billion, respectively. Nearly four out of every five dollars (79.2 percent) that Medicare and Medicaid spends caring for dual eligibles is paid for with federal funds: Collectively, in 2007 the federal government paid $210.4 billion on dual eligibles—$141.0 billion through Medicare and $69.4 billion through federal Medicaid matching payments; state Medicaid spending for the dual eligibles totaled $55.3 billion.

Figure 7

Medicare and Medicaid Spending Per Capita for Non-Elderly and Elderly Dual Eligibles, 2007

$16,422 $17,357

$11,325 $13,782

Dual Age <65 Dual Age 65+

Federal and StateMedicaid SpendingFederal MedicareSpending

Total = $27,747

Total = $31,139

Number of Beneficiaries = 5.6 million


SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured. Spending adjusted to 2007 NHE.

Figure 8

Total Medicare and Medicaid Spending for Dual Eligibles and Other Medicare Beneficiaries, 2007

$141.0

$256.0

$69.4

$55.3

Medicaid-Medicare DualBeneficiaries

Other Medicare Beneficiaries

State Medicaid Spending

Federal MedicaidSpendingFederal MedicareSpending


Total = $265.7 billionTotal = $256.0 billion



79.2% Federal

100% Federal

Figure 6

Medicare and Medicaid Spending Per Capita for Dual Eligibles and Other Medicare Beneficiaries, 2007

$15,850

$7,226

$14,018

Medicaid-Medicare DualBeneficiaries

Other Medicare Beneficiaries



Total = $29,868

Total = $7,226



006

MEDICAID AND MEDICARE’S SHARE OF SPENDING VARIES BY TYPE OF SERVICE

Although Medicaid and Medicare each finance about half of total public spending on dual eligibles, the distribution of spending by program varies considerably by type of service (Figure 9 and Table 2). Overall, Medicare paid the bulk (80.0 percent) of acute care services whereas Medicaid paid 20.0 percent. The share that Medicaid paid, however, ranged from a low of 2.8 percent for prescription drugs to a high of 79.6 percent for other acute care services, many of which are wrap-around services that Medicaid provides (some at state option) such as dental, vision and case management (Table 2). For sub-acute care, Medicare again finances the vast majority (81.7 percent) of these services with Medicaid paying 18.3 percent. However, for home health care (a subset of sub-acute care and for which Medicare’s coverage is limited), Medicaid paid 33.4 percent of these services. By contrast, Medicaid clearly dominates the financing of dual eligibles’ long-term care, both institutional and community-based care. For these services, Medicaid finances 100 percent of the spending.

These findings are consistent with the design of Medicare and Medicaid: Medicare covers many acute care services and Medicaid provides wrap-around coverage for those services by paying Medicare cost sharing for dual eligibles. Full dual eligibles also receive all Medicaid benefits, which include some additional acute and sub-acute care services beyond what Medicare covers (e.g., home health) and long-term care services not covered by Medicare. Of the total amount that Medicare spent on services for dual eligibles, 87.0 percent was for acute care services, with the remainder spent on sub-acute care (Table 2). Of the total amount Medicaid spent on services for dual eligibles, 72.1 percent was for long-term care, with 24.6 percent spent on acute care and 3.3 percent spent on sub-acute services (Table 2).

Spending patterns differ when examining spending among just those who use services (versus all dual eligibles) (Figure 10 and Table 2). While per capita spending figures provide an estimate of program expenditures for the dual eligible population, per user spending figures provide an estimate of the cost of services for a dual who requires a particular type of care. Comparing spending for dual eligibles on per capita and per user basis, we find little difference between the two for acute care services (Figure 10). This reflects the fact that virtually all dual eligibles use acute care services, particularly ambulatory care and prescription drugs. For sub-acute and long-term care, however, we find a substantial difference between the per capita and per user estimates, reflecting the more limited use of these services (Figure 10). The most expensive service, by far, is institutional long-term care, which was $42,410 per user (Table 2). While lower, per user costs for community-based long-term ($21,242) are also high (Table 2).

Figure 9

Medicare and Medicaid Spending for Dual Eligibles by Service Type, 2007

$122.6

$18.4

$30.6

$4.1

$89.9

Acute Care Sub-Acute Care Long-Term Care



Total = $153.2 billion



00 7

Figure 10

Medicare and Medicaid Spending Per Capita and Per User for Dual Eligibles by Service Type, 2007

$13,783 $14,329

$2,067

$11,625

$3,444 $4,368

$464

$7,924

$10,110

$33,689

Per Capita Per User Per Capita Per User Per Capita Per User

Federal Medicare Spending Federal and State Medicaid Spending


Sub-Acute Care Long-Term CareAcute Care

WHO ARE THE HIGH-COST DUAL ELIGIBLE BENEFICIARIES?

While on average dual eligibles are a high-cost group, there is considerable heterogeneity among them. As is the case for health spending generally, a relatively small portion of dual eligibles are responsible for the majority of the group’s total spending, with fewer than 20 percent of dual eligibles accounting for nearly 60 percent of combined Medicaid and Medicare spending for dual eligibles. Because of the split in the types of services covered by Medicare and Medicaid, what makes beneficiaries high-cost in one program does not necessarily make them high-cost in the other. Indeed, our analysis finds relatively small overlap in the highest spenders in the two programs (Figure 11). Looking at the top 10 percent of dual eligibles in terms of Medicaid spending—those with more than $45,180 in Medicaid spending—fewer than 15 percent are also in the top 10 percent of dual eligibles in Medicare spending—those with Medicare spending greater than $44,348. Only about 83,000 dual eligibles are in the highest 10 percent of the spending distribution in both Medicaid and Medicare.

Figure 11

Dual Eligible Enrollment and Spending by High-Cost Groups, 2007

7.2

$104.8

0.8

$67.8

0.1

$12.3

0.8

$80.8

Beneficiaries (millions) Spending (billions)

Top 10% of MedicareSpendersTop 10% of Spenders in BothMedicare and MedicaidTop 10% of MedicaidSpendersNot a Top 10% spender inEither Program

SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured. Spending adjusted to NHE.NOTE: Top 10% Medicaid spenders = Medicaid spending greater than $45,180 and Top 10% Medicare spenders = Medicare spending greater than $44,348.

Total = 8.9 million Total = $265.7 billion

008

Focusing on the approximately 900,000 dual eligibles who constitute the top 10 percent of Medicaid spenders (Figure 12; left pie), we find that 74 percent of their combined Medicare and Medicaid spending is for long-term care while just 23 percent is for acute care. Thus, the high costs for this group of dual eligibles are driven largely by spending on costly long-term care services, such as institutional care. In stark contrast, for the top 10 percent of Medicare spenders (Figure 12; right pie), 78 percent of their spending was for acute care services and less than 10 percent is for long-term care services. Thus the high costs for this group of dual eligibles are driven by intense use of acute care services such as inpatient hospital care.

An examination of the demographic and health characteristics of top spending dual eligibles for Medicaid to their lower-cost counterparts suggests several factors that distinguish the two groups (Figure 13). Compared to lower-cost Medicaid dual eligibles—that is, dual eligibles in the bottom half of the Medicaid spending distribution (or with Medicaid spending less than $1,182)— top dual Medicaid spenders are more likely to be over age 80, white, non-Hispanic, and less likely to be married. Living in an institution and, closely related, having 3 or more ADL limitations come close to completely determining who a high-cost dual for Medicaid is. Finally, top spending Medicaid dual eligibles are substantially more likely than lower-cost Medicaid dual eligibles to have diabetes or Alzheimer’s disease.

Figure 13

Selected Characteristics of Top 10% and Bottom 50% of Dual Eligible Medicaid Spenders, 2007

3.1

0.6

9.0

1.7

22.0

59.2

19.3

14.6

18.5

75.2

75.8

12.8

72.1

47.7

Alzheimers

Diabetes

3+ ADL

Living in an Institution

Married

White Non-Hispanic

Age > 80

Top 10% Medicaid Spenders Bottom 50% Medicaid Spenders

SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for the Kaiser Commission on Medicaid and the Uninsured. Spending adjusted to 2007 NHE.NOTE: Top 10% Medicaid spenders = Medicaid spending greater than $45,180; Bottom 50% Medicaid spenders = Medicaid spending less than $1,182. * (**) (***) Top 10% Medicaid spenders are significantly different from bottom 50% Medicaid spenders at the .10 (.05) (.01) level, two-tailed test.

***

***

***

***

***

***

***

Figure 12

Spending by Service Among High-Cost Duals, by Program, 2007

Acute Care,

$18.4B 23%

Sub-Acute Care, $2.8B

4%

Long-Term Care,

$58.9B 74%

Top 10% Medicaid Spenders (N=0.9 million)

Acute Care,

$73.1B 78%

Sub-Acute Care,

$11.7B 13%

Long-Term Care, $8.3B

9%

Top 10% Medicare Spenders (N=0.9 million)

SOURCE: Urban Institute analysis of MSIS-MCBS 2007 linked file for KCMU. Spending adjusted to NHE.NOTE: Top 10% Medicaid spenders = Medicaid spending greater than $45,180 and Top 10% Medicare spenders = Medicare spending greater than $44,348.

Total = $80.1 billion Total = $93.1 billion

00 9

CONCLUSIONS

On average, dual eligibles are among the most costly beneficiaries in both Medicare and Medicaid. They are generally both sicker and poorer than other Medicare beneficiaries and, correspondingly, use more services and cost more per capita. Medicare and Medicaid share the cost of caring for dual eligibles, with Medicaid financing just under half of the spending on the group. Compared to previous years,8 Medicaid’s share of cost for dual eligibles has declined somewhat, largely due to the 2006 implementation of a prescription drug benefit in Medicare that shifted responsibility for this service from Medicaid to Medicare.

Consistent with previous work, we find considerable diversity among dual eligibles in their demographic and diagnostic profiles and patterns of service use. We also find that being a dual eligible is not synonymous with high-spending. Some dual eligibles have relatively low service needs and corresponding cost, and some dual eligibles are eligible only for a limited set of benefits from Medicaid (specifically, assistance with Medicare cost sharing). Other dual eligibles, however, have very substantial health care needs. Many with high needs rely heavily on Medicare for acute care services, while others with high needs rely heavily on Medicaid for long-term care services. Only a small number of dual eligibles—less than a million—are heavy users of both Medicare acute care and Medicaid long-term care services.

The diversity in the dual eligible population is important to the development of effective strategies to finance their care and control costs. Importantly, the factors that make dual eligibles more expensive than other Medicare beneficiaries are not the same factors that make them expensive relative to other Medicaid beneficiaries. For Medicare, the reason for high costs among dual eligibles is the elevated need for acute care resulting from the increased prevalence of chronic disease associated with age, disability and poverty. But for Medicaid, the principal reason that dual eligibles tend to be expensive is that they are more likely than other dual beneficiaries to be users of institutional long-term care.

These results suggest that decision-makers should adopt a multi-pronged approach in developing approaches to improve care delivery and efficiency for dual eligibles. This is challenging in itself, and with the split in financing obligations between Medicare and Medicaid the challenge becomes that much greater. On one hand, the clear distinction in responsibilities could enable each program to focus on the services most relevant to its costs. On the other hand, if each program focuses only on its own high-cost users in its efforts to address coordination/integration, they will be missing most of the dual eligibles who are of importance to the other program. Further, uncoordinated policy action between the two programs could have offsetting effects: for example, Medicare efforts to address high costs for sub-acute care could lead to higher Medicaid spending for sub-acute or even long-term care. Recent initiatives to address fragmentation in care delivery for dual eligibles, including the Center for Medicare and Medicaid Innovation and the Medicare-Medicaid Coordination Office within the Centers for Medicare and Medicaid Services (CMS), aim to integrate benefits and align financing among both programs. These efforts hold potential to help the two programs work together to address cost and quality but need to recognize the diversity in the dual eligible population that may call for multiple approaches.

0010

This brief was prepared by Teresa A. Coughlin, Timothy Waidmann, and Lokendra Phadera of The Urban Institute and Rachel Garfield and Barbara Lyons of the Kaiser Commission on Medicaid and the Uninsured. An accompanying article drawing on these findings appears in an April 18, 2012 Health Affairs Web Exclusive.

00 11

Table 1Selected Characteristics of Medicaid-Medicare Dual Eligibles and Other Medicare Beneficiaries, 2007

All Medicare Beneficiaries Dual Eligibles

All Dual Non-dual < 65 65 + (column %) (column %) (column %) (column %) (column %)

N 44,327,346 8,896,020 35,431,326 *** 3,333,854 5,562,166 ###

Dual Status

Full Medicaid benefits 12.1 76.3 NA 80.6 73.6 ###

Medicare Savings Plan 3.6 22.5 NA 18.7 24.8 ###Unknown 0.2 1.2 NA 0.7 1.5 ##Non-Dual 84.1 NA 100.0 NA NA

Functional StatusNo ADLs 77.8 56.7 81.8 *** 66.5 50.9 ###1-2 ADL 12.9 18.1 11.9 *** 20.4 16.8 #3 + ADL 9.3 25.1 6.3 *** 13.1 32.3 ###

Living ArrangementCommunity 94.3 79.2 97.1 *** 87.9 74.0 ###Institution 4.3 17.0 1.9 *** 9.3 21.7 ###Both 1.4 3.8 1.0 *** 2.8 4.3 #

Age < 65 16.0 37.5 11.9 *** 100.0 NA65 and over 84.0 62.5 88.1 *** NA 100.0

GenderMale 44.6 35.7 46.3 *** 48.1 28.3 ###Female 55.4 64.3 53.7 *** 51.9 71.7 ###

EthnicityWhite non-Hispanic 79.4 59.5 83.2 *** 61.7 58.3Black non-Hispanic 9.2 18.6 7.4 *** 22.3 16.4 ###Hispanic 6.5 13.1 5.3 *** 9.6 15.2 ###Other 4.9 8.8 4.2 *** 6.5 10.1 ##

Family StructureMarried 52.0 20.4 57.9 *** 15.8 23.2 ###Single/Divorced/Separated/Widowed 47.6 79.4 41.6 *** 83.8 76.8 ###

Income$10,000 or less 15.4 53.4 8.3 *** 58.1 50.5 ###$10,001 - 20,000 27.2 39.1 24.9 *** 35.0 41.6 ###$20,001 or more 57.4 7.5 66.8 *** 6.9 7.9

SOURCE: MSIS-MCBS 2007 linked file.Weighted.NOTE: NA: Not applicable. Totals may not sum to 100 percent due to rounding and exclusion of missing values.

*(**)(***) Medicaid-Medicare Dual beneficiaries are significantly different from other Medicare beneficiaries at the .10 (.05) (.01) level, two-tailed test.#(##)(###) Non-elderly dual beneficiaries are significantly different from elderly dual beneficiaries at the .10 (.05) (.01) level, two-tailed test.

0012

Tabl

e 2

Serv

ice

Use

and

Spe

ndin

g Pa

ttern

s by

Dua

l Elig

ible

s, 2

007

Per C

apita

Spe

ndin

g ($

)Pe

r Use

r Spe

ndin

g ($

)To

tal S

pend

ing

($ m

illio

ns)

Com

bine

dM

edic

are

Med

icai

dC

ombi

ned

Med

icar

eM

edic

aid

Com

bine

dM

edic

are

Med

icai

dSh

are

Med

icai

d

Tota

l Spe

ndin

g29

,868

15,8

5014

,018

30,3

4716

,474

17,3

2326

5,70

214

1,00

012

4,70

346

.9%

Acu

te C

are

17,2

2713

,783

3,44

417

,504

14,3

294,

368

153,

252

122,

615

30,6

3720

.0%

Inpa

tient

Car

e8,

312

7,86

444

826

,839

27,8

154,

017

73,9

4569

,957

3,98

85.

4%A

mbu

lato

ry C

are

3,92

82,

629

1,29

94,

155

2,86

32,

156

34,9

4223

,384

11,5

5833

.1%

Pre

scrip

tion

Dru

gs2,

961

2,87

883

3,39

03,

738

224

26,3

4225

,601

741

2.8%

Oth

er A

cute

Car

e (in

clud

es c

apita

tion)

2,02

641

31,

613

2,62

21,

038

2,39

818

,024

3,67

314

,350

79.6

%

Sub-

Acu

te C

are

2,53

12,

067

464

11,7

6311

,625

7,92

422

,514

18,3

854,

129

18.3

%S

kille

d N

ursi

ng F

acilit

y C

are

1,13

91,

139

NA

12,6

8612

,686

NA

10,1

3310

,133

NA

0.0%

Hom

e H

ealth

Car

e1,

392

928

464

8,56

17,

584

7,92

412

,381

8,25

24,

129

33.4

%

Long

-Ter

m C

are

10,1

10N

A10

,110

33,6

89N

A33

,689

89,9

36N

A89

,936

100.

0%In

stitu

tiona

l Car

e6,

789

NA

6,78

942

,410

NA

42,4

1060

,393

NA

60,3

9310

0.0%

Com

mun

ity-B

ased

Car

e3,

321

NA

3,32

121

,242

NA

21,2

4229

,543

NA

29,5

4310

0.0%

Sou

rce:

MS

IS-M

CB

S 2

007

linke

d fil

e. M

edic

are

spen

ding

adj

uste

d to

200

7 N

atio

nal H

ealth

Exp

endi

ture

Acc

ount

s (N

HE

).N

otes

:Wei

ghte

d.

NA

: Not

app

licab

le.

Med

icar

e am

bula

tory

ser

vice

s in

clud

e M

edic

are

phys

icia

n an

d ou

tpat

ient

ser

vice

s. M

edic

aid

ambu

lato

ry s

ervi

ces

incl

ude

clin

ic, m

id-w

ife, n

urse

pra

ctiti

oner

, oth

er p

ract

ition

er

(chi

ropr

acto

rs, p

odia

trist

s, p

sych

olog

ist,

and

opto

met

rists

), ou

tpat

ient

, and

phy

sici

an s

ervi

ces.

Med

icar

e ot

her a

cute

ser

vice

s in

clud

es h

ospi

ce, d

urab

le m

edic

al e

quip

men

t and

cap

itatio

n pa

ymen

ts. M

edic

aid

othe

r acu

te s

ervi

ces

incl

udes

hos

pice

, tra

nspo

rtat

ion,

ther

apy

(spe

ech/

lang

uage

, occ

upat

ions

, phy

sica

l), ta

rget

ed c

ase

man

agem

ent,

relig

ious

non

-med

ical

, reh

abilit

atio

n, p

rivat

e-du

ty n

ursi

ng, p

rimar

y ca

re c

ase

man

agem

ent,

pre-

paid

he

alth

pla

ns, l

ab/x

-ray

, HM

O, d

enta

l, ab

ortio

n an

d ot

her s

ervi

ces

such

as

pros

thet

ics

and

eyeg

lass

es.

M

edic

aid

long

-term

inst

itutio

nals

ervi

ces

incl

udes

nur

sing

faci

litie

s, in

term

edia

te c

are

faci

litie

s fo

r men

tally

reta

rded

(IC

F-M

R),

and

IMD

ser

vice

s.

M

edic

aid

long

-term

com

mun

ity-b

ased

ser

vice

sin

clud

epe

rson

al c

are

and

hom

e an

d co

mm

unity

-bas

edse

rvic

es (H

CB

S).

00 13

APPENDIX Data For this study, we relied on two data sources: 1) the 2007 Medicare Current Beneficiary Survey (MCBS) that has been adjusted to reflect National Health Expenditures and 2) the 2007 Medicaid Statistical Information System (MSIS) Summary File that has been adjusted to Medicaid spending as reported in the CMS-64. The MCBS provides detailed information on spending by Medicare and Medicaid, as well as spending by other payment sources such as private insurance or beneficiary out-of-pocket spending. In this study, however, we focus on spending by public programs. As such, our analysis examines only spending for Medicare beneficiaries that is paid by Medicare, and for dual eligibles only spending paid by Medicare and by Medicaid. Limiting spending to what is paid for by Medicare and Medicaid explains why we find that other Medicare beneficiaries (non-dual eligibles) have no long-term care spending. If we included spending by other sources (e.g., out-of-pocket or private LTC insurance) other Medicare beneficiaries would show spending for LTC services, as well as higher spending on other services. MCBS. The MCBS is a survey of a nationally representative sample of Medicare beneficiaries. It is conducted annually by the Centers for Medicare & Medicaid Services (CMS)9 and collects a wealth of information, including expenditures and sources of payment for all services used by Medicare beneficiaries, as well as the types of health insurance coverage beneficiaries had over the course of the year. For this analysis we used the 2007 MCBS Cost and Use (CU) file, which provides information on Medicare beneficiaries who were ever enrolled in the program during the calendar year. This includes those who died before the end of the year, as well as those who enrolled during the year. The CU file, which is a combination of MCBS survey and Medicare administrative data, contains information on several health care services used and their associated costs, as well as a wide variety of information on health status, social and demographic characteristics. It also includes information on prescription drugs and long term care (LTC) services. To account for the problem of underreporting of medical service use associated with personal interview surveys10, MCBS survey data are matched with more accurate administrative bill data, and missing or under/over-reported information are filled-in if not matched. For this study, we used several data elements from the 2007 CU file, including demographics, living arrangement, income, functional status, and Medicare spending. While the MCBS collects information on Medicaid spending, as we will discuss below, we relied on the MSIS for Medicaid spending information. Although the MCBS contains self-reported data on Medicaid spending, it does not provide accurate spending information for home and community-based waiver programs, an important expenditure item for the dual population. For this reason we relied exclusively on MSIS for Medicaid spending information for the study. In our analysis we found that the MCBS underestimated the national estimates of personal health care expenditures by Medicare. To address this problem, we matched our 2007 Medicare spending estimates to the 2007 National Health Expenditure (NHE) accounts by type of service. Specifically, we took the ratio between our estimate and NHE’s record, in particular the Personal Health Care Expenditures

0014

account, for each service and used it to inflate or deflate Medicare spending on that service by both dual eligibles and non-dual eligibles. Importantly, even with the adjustment to the NHE, our Medicare spending estimates do not include Medicare special payments such as hospital disproportionate share (DSH) or graduate medical education payments. MSIS. For Medicaid information we relied on the 2007 Medicaid Statistical Information System (MSIS) Summary File, which covers the 2007 federal fiscal year—that is, October 1, 2006 to September 30, 2007. Under federal Medicaid law, all 50 states and the District of Columbia are required to submit Medicaid eligibility and claims data to the Centers for Medicare and Medicaid Services (CMS) on a quarterly basis. Once received by the CMS, the data are subject to quality assurance edits and validity checks. For each federal fiscal year, the MSIS Summary File contains individual level data with aggregate measures of Medicaid expenditures for 30 service categories, including ambulatory, acute and institutional services, for all persons who received Medicaid services nationwide. For complete Medicaid service expenditures, we matched the MSIS service expenditures to those reported in the CMS-64. The CMS-64 is considered a more accurate financial data source for Medicaid service spending than MSIS because states use CMS-64 data to receive their federal matching payments. By benchmarking to the CMS-64, our expenditure estimates include special payments made by Medicaid such as hospital DSH payments.11 Expenditures reported in MSIS include payments made to providers on behalf of dual eligibles for Medicare cost-sharing but they do not include payments to Medicare for premiums. To account for these payments, we calculate premiums paid by Medicaid to Medicare for each acute and sub-acute care services, reduce those amounts from their respective spending categories in Medicare, and add them to Medicaid spending categories. Also, for dual eligibles enrolled in managed care, MSIS only includes capitation payments. No information on spending by type of service is available for dual eligibles in capitated managed care. While a limitation, less than 12 percent of dual eligibles were in enrolled in capitated Medicaid managed care arrangements in 2009.12 In addition to expenditure data, the MSIS Summary File contains information on the personal characteristics of Medicaid beneficiaries such as date of birth, sex, and race/ethnicity, as well as state and county of residence. MSIS also includes eligibility information that specifies why an individual is enrolled in Medicaid. We know, for example, if dual individuals are enrolled because they are disabled or aged. To identify whether the beneficiary was a full-benefit dual or a Medicare Savings Plan beneficiary, we used the MSIS variable that describes a beneficiary’s most recent or “last-best" monthly eligibility code. Linking the MCBS with the MSIS. Using the MCBS as the base file, we linked it to the MSIS to derive our sample of individuals enrolled in both Medicare and Medicaid in 2007. We used a crosswalk of unique identifiers in each of the two datasets that was provided to us by CMS. Specifically, we used the MCBS BaseID, the MSIS_ID, gender, and date of birth (allowing a difference of 31 days), for the match. Due to unavailability of Medicaid spending data for US territories other than the 50 states and the District of Columbia, of the 11,995 Medicare beneficiaries in the 2007 MCBS CU file, 207 beneficiaries from Puerto Rico were excluded. The CMS crosswalk file included both full benefit dual eligibles and enrollees in Medicare Savings Programs, sometimes referred to as “partial dual eligibles.” After identifying potential dual eligibles, we excluded cases that did not match on the crosswalk variables in both the MCBS and MSIS. Further, some dual beneficiaries had Medicaid records in more than one state in the MSIS. In these cases, we combined a beneficiary’s spending across states and

00 15

reported as a single record. The final number of dual eligibles from the linked MCBS-MSIS file we used in the analysis was 2,531. As mentioned, the MCBS captures spending in a particular calendar year whereas MSIS captures spending in a federal fiscal year, which runs October 1 to September 30. Thus while we analyze 12 calendar months’ worth of spending from each dataset, the timing of spending information between the programs is off by three months with some of the Medicaid spending being in the last quarter of 2006. While an admitted shortcoming, the study data still provide relative estimates of Medicare and Medicaid spending for dual beneficiaries over a given time period. Methods Sub-acute and long-term care services are divided into institutional and community-based care. Institutional care is provided by nursing facilities, intermediate care facilities for the mentally retarded (ICF-MR) and institutions for mental disease (IMD). Because of the fundamental difference in the nature and duration of these benefits we separate the Medicare and Medicaid covered services in our calculations.13 Community based care is divided into home health services, covered by both Medicaid and Medicare, personal care services (Medicaid only), and long-term care provided under Medicaid’s home and community based service (HCBS) waiver programs. All analyses presented are weighted using the MCBS cross-sectional weight.14 While this weight is intended for use in cross-sectional statistics involving the total (combined) national sample, it also can be used for analyzing representative subgroups. Given the relatively high match rate between the MCBS and the MSIS, we feel the combined data file is a representative sample of the national dual eligible population.

NOTES 1 Rousseau, D, L Clemans-Cope, E Lawton, J Langston, J Connolly, and J Howard. 2010 “Dual Eligibles: Medicaid Enrollment and Spending for Medicare Beneficiaries in 2007.” Kaiser Commission on Medicaid and the Uninsured, Washington, D.C., December. Kaiser Family Foundation. 2011 “The Role of Medicare for People Dually Eligible for Medicare and Medicaid.” Washington, DC, January. 2 Kaiser Commission on Medicaid and the Uninsured, “Proposed Models to Integrate Medicare and Medicaid Benefits for Dual Eligibles: A Look at the 15 State Design Contracts Funded by CMS,” Aug., 2011, available at http://www.kff.org/Medicaid/8215.cfm; Kaiser Commission on Medicaid and the Uninsured, “Financial Alignment Models for Dual Eligibles: An Update,” Nov., 2011, available at http://www.kff.org/medicaid/8260.cfm 3 Coughlin, TA, T Waidmann, and MO Watts. 2009. “Where Does the Burden Lie? Medicaid and Medicare Spending for Dual Eligible Beneficiaries.” Kaiser Commission on Medicaid and the Uninsured, Washington, D.C., April. 4 A detailed explanation of the data used in the analysis is provided in the Appendix. 5 Kaiser Commission on Medicaid and Uninsured. 2010. “Dual Eligibles: Medicaid’s Role for Low-Income Beneficiaries.” Washington D.C., December. 6 Health conditions examined were diabetes, heart disease, COPD/lung disease, mental illness, Alzheimer’s, and mental retardation. 7 Utilization measures only services paid for by Medicare or Medicaid. Health care services paid for private insurance or out of pocket by individuals are not included.

0016

8 Coughlin, TA, T Waidmann, and MO Watts. 2009. “Where Does the Burden Lie? Medicaid and Medicare Spending for Dual Eligible Beneficiaries.” Kaiser Commission on Medicaid and the Uninsured, Washington, D.C., April. 9 For more information on the MCBS see http://www.cms.gov/LimitedDataSets/11_MCBS.asp. 10 Survey respondents may forget some medical services used and might find it hard to recall those events during interviews, especially if the incidents are minor and for recall periods that are very long. Given the complex payment system the US health care system has, it might be even harder for the respondents to recall expenditures for medical services used. For more on the matching of MCBS survey data with Medicare claims see Eppig, Franklin J. and Brad Edwards. “Computer Matching of Medicare Current Beneficiary Survey Data with Medicare Claims.” IN. Richard Warneck, ed. “Health Survey Research Methods: Conference Proceedings.” DHHS Publication Number 96-1013. National Center for Health Statistics, U.S. Department of Health and Human Services, Hyattsville, MD. 1996. 11 Thus, while Medicaid spending estimates are adjusted for special program payments, as mentioned, Medicare spending estimates are not, an admitted shortcoming. 12 Centers for Medicare & Medicaid Services (CMS). 2009. “2009 Medicaid Managed Care Enrollment Report.” Washington, D.C., June. 13 Medicare does not cover long term care, but does pay for institutional and community based care after hospital stays. Institutional post-acute care is covered for a limited duration by the Medicaid skilled nursing facility benefit while community based care for home-bound individuals is covered by the Medicare home health benefit. Medicaid covers a wider range of long-term care services. 14 Based on the number of dual eligibles identified in the MSIS, the MCBS cross-sectional weight underestimates number of dual eligibles and hence services use and spending by them. As a result, number of dual eligibles estimated is 26 percent less than the actual administrative records. Some of this difference can be made up by including so-called “ghost” records from the MCBS – donor records from other sample members intended to make the MCBS population and spending estimates match administrative totals. Tabulations of demographic characteristics and per capita spending among the dual-eligible “ghosts” finds them to be quite similar to the matched sample. Assuming that the missing/non-estimable dual eligibles have same characteristics and service use patterns as the non-missing/estimable dual eligibles, we inflate our estimates for total number of dual eligibles and their total spending by 26 percent.

1 3 3 0 G S T R E E T N W , W A S H I N G T O N , D C 2 0 0 0 5

P H O N E : ( 2 0 2 ) 3 4 7 - 5 2 7 0 , F A X : ( 2 0 2 ) 3 4 7 - 5 2 7 4

W E B S I T E : W W W . K F F . O R G / K C M U

A d d i t i o n a l c o p i e s o f t h i s r e p o r t ( # 0 0 0 0 ) a r e a v a i l a b l e o n t h e K a i s e r F a m i l y F o u n d a t i o n ’ s w e b s i t e a t w w w . k f f . o r g .

The Kaiser Commission on Medicaid and the Uninsured provides information and analysis on health care coverage and access for the low-

income population, with a special focus on Medicaid’s role and coverage of the uninsured. Begun in 1991 and based in the Kaiser Family

Foundation's Washington, DC office, the Commission is the largest operating program of the Foundation. The Commission's work is

conducted by Foundation staff under the guidance of a bi-partisan group of national leaders and experts in health care and public policy.

This publication (#7895-02) is available on the Kaiser Family Foundation’s website at www.kff.org.

the diversity pf dual eligible beneficiaries: an ... · pdf filemedicaid kaiser i commission...

Documents